PROJECT SUMMARY Subthalamic deep brain stimulation (DBS) is an effective for the treatment of Parkinson's disease (PD). However, little is known about the specific neural pathway(s) responsible for therapeutic benefit. Direct stimulation of either the hyperdirect pathway and/or the subthalamopallidal pathway represents two of the most likely candidates as the target. Therefore, the goal of this project is to combine patient-specific 7T imaging and neurostimulation models together to enable probabilistic identification of the stimulation pathways linked to changes in clinical outcome measures recorded from subthalamic DBS patients. This study will rely on tractography-activation models (TAMs) to guide clinical testing on a cohort of 30 subthalamic DBS patients. Our multi-disciplinary approach will integrate the latest advances in neuroimaging, neurostimulation modeling, and quantitative clinical outcome measures to customize DBS to these patients. The first aim will acquire high- resolution 7T MRI data prior to their implant surgery. These images will provide unparalleled anatomical characterization of each patient. The second aim will create patient-specific TAMs for each subject. These models will enable us to define theoretically optimal settings for focused activation of either the hyperdirect pathway or the subthalamopallidal pathway. The third aim will quantify the clinical outcomes achieved with stimulation parameters defined by either the TAM or traditional clinical practice. These clinical results will be coupled with the TAM predictions to create a probabilistic tractography-activation atlas (PTAA) for the subthalamic region. The results of this study will help identify optimal implantation locations for DBS electrodes, and enable theoretical prediction of stimulation parameter settings that focus activation on the targeted pathways and/or avoid side-effect pathways.